94 
SIR W. DE W. ABNEY AND PROF. W. WATSON ON 
The amount by which the light coming through the slit in the spectrum is reduced 
owing to its passage through the annulus depends for any given position of the annulus 
not only on the colour of the light but also on the distance of the receiving surface, I, 
from the annulus. The reason is that in the first place the coefficient of absorption 
of the annulus is a function of the wave-length. Secondly, there is a certain small 
amount of light scattered by the annulus, and the amount of this scattered light which 
falls on the disc depends on the distance. Hence it was necessary to measure the 
absorption of the annulus throughout the spectrum in exactly the same relative 
position as that in which it is used. This calibration involved a very considerable 
amount of work but need not be described. 
It enabled us to calculate for any wave-length and for any annulus reading what 
fraction of the light passing through the slit fell on the disc. The intensity of the 
light falling on the disc, when the slit was placed at the sodium line, was determined 
by comparison with a standard Hefner lamp. This lamp is very suitable for the 
purpose, as the colour of its light does not differ greatly from that of the D line. 
The procedure adopted was for the observer, who had been in complete darkness 
for more than half-an-hour, to look steadily with both eyes at the fixation spot 
and then to press the key, M, for about a second, thus allowing the light to fall on the 
disc, I, for this time. If the light was visible, he then turned the annulus so as 
to reduce the light, and again pressed the key. This process was continued till the 
light was just imperceptible when the key was pressed. The circle attached to the 
annulus having been read, the annulus was turned so as to make the light brighter, 
when the setting was repeated. In general, three settings were taken at each 
selected wave-length. 
When observing away from the fovea the fixation spot was always placed vertically 
above the disc on which the stimulus light was received. Unless otherwise mentioned, 
the disc, on which the light was received, had a diameter of 6‘3 mm., and was at 
a distance of 72 cm. from the observer’s eye, so that the angle subtended at the eye 
was 34 minutes. 
With the fixation spot at 1 '5 mm. or more away from the centre of the disc, all the 
observers found it quite easy to obtain consistent results. With foveal fixation, some 
observers required a little practice before they were able to overcome the tendency for 
the eye to wander. The reason for this tendency is that for most people the sensi¬ 
tiveness of the retina increases (except in the red), that is the threshold value 
decreases, rapidly as we go out from the fovea, and hence, when striving to see the 
last glimmer of light they instinctively shift the axis of the eye so as to bring the 
image on the more sensitive area surrounding the fovea. The curves shown in 
fig. 2 illustrate the successive readings of the annulus obtained by one untrained 
observer with central fixation.* At the first attempt the numbers obtained correspond 
* The numbers plotted in this figure are the annulus readings. The greater the annulus reading the 
greater the reduction in the light. 
